Projects per year
Seeds are transported by flowing water along with sediment and organic matter and form seed banks within the fluvial landforms (or geomorphic units) of river systems. Fluvial processes commonly result in observable gradients in sediment characteristics with elevation above the channel bed. This study aims to investigate spatial variability in seed bank characteristics that may be attributed to fluvial processes. We compared the extent to which four riparian seed bank characteristics (abundance, species richness, seed mass, and seed shape) are correlated with (1) fluvially influenced gradients in sediment character (percentage of organic matter [organic %], median particle size [D50], percentage of fine particles [% fines], and the combined percentage of sand and gravel [% sand/gravel]); and (2) three geomorphic unit types that form with increasing elevation above the channel bed: bars, benches, and the floodplain. Seed bank abundance and species richness were weakly yet significantly correlated with sediment gradients, significantly increasing with increasing % fines and decreasing % sand/gravel. Seed bank species richness also significantly increased with increasing organic % and decreasing D50. Conversely, seed abundance was highly variable, and relationships between sediment qualities and seed mass and shape were nonexistent. We suggest that hydrological factors such as inundation frequency, operating most clearly at the geomorphic unit scale, ultimately drive spatial variability in riparian seed bank characteristics by (1) directly influencing seed inputs and losses from seed banks via erosion and deposition, which is complicated by species-specific differences in seed morphology; and (2) indirectly influencing seed inputs and losses by mediating seed germination and mortality and the establishment success of plants, through impacts on sediment moisture and organic matter content. Further complexity is added by nonfluvial seed inputs from established vegetation. The net result of these influences tends toward decreasing potential for seed losses and increasing potential for seed inputs associated with the reduction in inundation frequency observed from bars to the floodplain. For those assessing the potential of the seed bank as a seed source for revegetation, sediment sampling for seed bank assays targeted at organic flood debris, or fine, organic-rich sediments such as those typically associated with benches and the floodplain, should elicit useful estimations of seed bank composition.